This study is designed to determine whether the statistical technique of blocking on baseline DMFS should be used in the design and analysis of dental clinical trials, and to provide specific recommendation for its implementation if it is judged desirable. In addition to treating baseline DMFS as a blocking factor, baseline DMFS is studied as both a single covariate, and also as separate covariates for each group. The problem of lack of balance in blocks and/or groups will be studied.